# Predicting human mobility with activity changes

**Repository Path**: tjcsgi/predicting-human-mobility-with-activity-changes

## Basic Information

- **Project Name**: Predicting human mobility with activity changes
- **Description**: Implementation of Predicting human mobility with activity changes.

- **Primary Language**: Unknown
- **License**: MIT
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2021-12-14
- **Last Updated**: 2021-12-14

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# Predicting human mobility with activity changes

Implementation of [Predicting human mobility with activity changes](https://www.tandfonline.com/doi/abs/10.1080/13658816.2015.1033421)

## Usage
**Please read the source code before you execute python scripts.** Some configuration need to change in the file for custom.
### Extraction of human activity
You need to put MSRA Geolife dataset `Data` folder in the same level directory of the project.

```shell
python extra_stay_pints.py
```

### Activity clustering
We provide two method to cluster, you can read `dbscan_cluster.py` for details.
```shell
python cluster.py
```

### Human activity change detection
For threshold $d$, here we provide several options and you can try for best performance.
```shell
python change_detection.py
```

### Prediction of human movement
The script just as an auxiliary for prediction, and you need to do prediction on your own method.
```shell
python markov.py
```

## Reference
[1] Wei Huang, Songnian Li, Xintao Liu & Yifang Ban (2015) Predicting human mobility with activity changes, International Journal of Geographical Information Science, 29:9, 1569-1587, DOI: 10.1080/13658816.2015.1033421